For a still-contemporary comparison, Trader Joe’s still publishes a similar (but more condensed and less personal) story-per-item format catalog for its grocery offerings in its Fearless Flyer, which you can find at https://www.traderjoes.com/home/ff

The collective suffering of working-class commuters ascending and descending the income gradient on packed freeways is the thing that powers LA, like children’s screams powered the city in Monsters, Inc. Robotaxis can never replace that.

From what I learned about it since that time, the details John shares in the book are definitely much more significant factors in why things happened why they did. And if anyone needed to learn what the words “vertical slice” meant, it was the DK production team. Programming the sidekicks, a definitional feature, was left until close to the end of the project, with disappointing results.

So the complaints about how hard staffing was and how people didn’t want to come to Dallas were straight from John’s mouth, but I think it had a lot to do with the state of the project and Ion. Steve Ash (RIP), our fourth lead programmer, was just about ready to return to California where he would end up helping to start Double Fine, so that situation was on his mind (speaking to me as a California fly-in AI programmer).

But as the 1300x960 arrow story typified, experienced developers were hard to find as team sizes were doubling from 20 to 40 throughout the industry. At the same time, Daikatana was being roasted constantly on Old Man Murray, Something Awful and various messageboards, and Half-Life made Daikatana’s story and cinematic ambitions seem less impressive. So by 1999, it’s fair to say Ion Storm was a hard sell as a place to work for a lot more reasons than the Dallas area...

* - https://www.dallasobserver.com/news/stormy-weather-6427649

To make this more HN-relevant, I will say that the whole point of Ion Storm was supposed to be unleashing Romero’s game direction, but the business partners who were supposed to give him that space instead provided another level of distraction.

John worked insanely hard and doesn’t blame others for what he’s responsible for, but with all the business chaos at Ion which he dealt with personally, he just could not be on top of everything and that’s the major reason why DK was not the epic game it could have been. That’s the part that he doesn’t want to say, but to me it’s clear.

It’s a good example of why a startup needs its product visionary highly focused at the most critical times. As a rough approximation, every night John went to bed thinking about Ion’s latest issues instead of thinking about Daikatana was a lost chance to make the game 2% better. That adds up.

  What is Brauer's theorem?
    [I had no idea and they moved on]

Since then, computer science has certainly been growing: “The number of students nationwide seeking four-year degrees in computer and information sciences and related fields shot up 34 percent from 2017 to 2022, to about 573,000, according to the National Student Clearinghouse Research Center.” https://www.washingtonpost.com/education/2023/05/19/college-...

Even at that rate, colleges cannot meet demand and CS majors are difficult to be admitted to even for the most competitive college applicants: https://www.geekwire.com/2022/it-is-not-acceptable-uw-comput...

The major is so sought-after that as of last year, UIUC has closed computer science for transfer from other majors: https://cs.illinois.edu/admissions/undergraduate/transfer-st...

It was mostly the same pitch that the showrunner gave to get the green light for the project, and it was crackling with energy, excitement and potential. It was magical.

So many other times, I was kind of tossed into a team and had to piece together what we were making and even why and how. This time, though, I went from knowing almost nothing about this project to having a clear view of where we wanted to go and being thrilled to be on this journey.

Speaking of which, I can’t resist sharing a fantastic metaphor of a creative project as a journey from Pixar director Andrew Stanton (the following are all his words):

The hardest thing about directing an animated movie is keeping yourself excited about it. It's hard enough to make the crew excited about it, but keeping yourself excited about it -- trying to remind yourself why you wanted to do it. Because it's all about the details once you really start making the movie.

It's no different than building a house, or, building a really extravagant mansion. There's a million details that you have to spend more time with after the bigger ideas of where the rooms are going to go and how it's going to be structured, and it can get you kind of bogged down.

Joe Ranft used to have this great expression that there's always a point during the making of a movie where there's sort of the Columbus where-is-the-land moment, where everybody on the boat is going "You promised us the land. Where's the land? We're not seeing it!" And people get bogged down in all the minor problems or the major problems that won't go away, and it's all justified -- it's all legitimate to have that response.

So, for me, to prevent that is to get really, really picky about what story you're going to tell up front. And this is my opinion, and it's not a rule. But if I have an idea that I kinda like, then I don't want to do it. If I have an idea that affects every fiber of my being, like "I want to see that movie made whether I make it or not" -- it's like that idea has to get on the screen -- that's a real good quality to start with. Because it's going to get attacked for the next four years. And there's going to be, sometimes, weeks or months where nothing seems to be going right.

[...] It's like looking for oil or something -- it's like "where can I find something that has enough fuel that's going to keep me going for years?" Because there's going to large stretches of time where nothing is working, nobody's happy, everybody thinks that the sky is going to fall, and what's going to get me out of bed is just because that idea still has to be on the screen.

So I want that when I'm going to go into battle. Because it's going to be battle. So if I don't have that going in, then I won't go into it -- I won't make that movie.

My idea was, wow, wouldn’t it be great to give other people this kind of expansive knowledge about Los Angeles? If they could only understand what I see on this map!

But of course, it’s not necessary to understand the entirety of a map if you have a tool that zeroes in on what you really want. Google doesn’t exist for you to understand the whole web, but to mine it precisely. Finding a home is hard to do the same way, but giving people a zeitgeist is always inferior to giving them a tool that lets them understand the least about something. It’s a hard lesson for an infovore to learn, I’ll say that.

So, of course, now LA drivers know too much, while they themselves know very little. In the 2000s there were “hidden shortcuts” and alternate routes that were risky. Now, the risk is hedged and nothing is hidden. It is a marketplace of nearly perfect information, where saving a minute of someone’s life in traffic is treated with the respect it actually deserves.

Yet now the big problem is that people have locked in their commute route years ago, and it gets slower and slower with no escape. They drive from their far-away apartment, guided by synthesized voices, past the landed aristocracy of Los Angeles with the favor of Prop. 13 whose lawn signs chide land-use refugees to drive like their non-existent kids lived there.

Ultimately, the era of broad knowledge was also an era of choice. Both are leaving us at the same velocity. There is little need for a giant book of colleges when you can’t get into or afford the ones you’re interested in. Who cares about a directory of reviewed doctors when your crummy insurance will not let you see any of the ones you would choose? The time of the Thomas Bros. map was before the closing of the urban frontier. Now, you just take the only thing on the shelf.

https://www.youtube.com/watch?v=FKyWNEGh_yk&t=30s

It's hard to describe exactly, but the technological "rules" about what experiences a video game could provide were not at all clear at that time, and there was still a mystique about the potential of every new video game.

Cube Quest had cutting-edge computer-generated animation (on video) and vector graphics, but to have Ken Nordine's amazing narration specifically recorded for the game was a new level in game production.

Ultimately, the spatial world inside the game (a tube shooter) was disappointing after all of the very cool presentation it was wrapped in. But before you went to wait in line for Atari's Star Wars, you would want to stick around again for the introduction.

NB: The on-screen instructions for Cube Quest may be the first documented use of the phrase "game play" for a video game.

But the "render just one frame" just means that package is a cut-down version of the "Animation" package which lets you render the entire scene while dispensing with the animation entirely.

It says that it doesn't explain the "conundrum" of increased higher-end scores. But it definitely addresses that it's very unlikely the test has gotten easier, since average scores have remained stable, as they have for the PSAT and ACT.

I don't think this is right. There is good evidence that the highest percentiles of students (the ones populating the competitive colleges mentioned) are indeed smarter, because they have standardized test scores to match.

First off, the bar for a PSAT score that gets National Merit recognition has risen significantly for most students in recent years. More graphically though, the number of students who get a 36 on the ACT goes back a while and is a good representation of the upper score band:

https://en.wikipedia.org/wiki/ACT_(test)#Highest_score

And indeed at the graduate level, GMAT scores at the top business schools are rising steadily:

https://www.usnews.com/education/best-graduate-schools/top-b...

So instead of the positive, hopeful story "our best students are becoming even more capable, year after year," we get a story like this shaming colleges and universities for "grade inflation" instead. Disappointing.

https://en.wikipedia.org/wiki/Luxury_tax_(sports)

I guess the real difficulty would be measuring how high San Francisco's housing-supply border wall is.

More relevantly, I would be surprised if the shift to AI techniques in fraud detection at places like PayPal is not already having an impact on the career paths of the engineers that were tasked with maintaining and tuning their pre-ML fraud system. At one point the top engineers of the original heuristic system could have been considered their most valuable non-management employees at the company. I'm sure they're not out on the streets or anything, but I also assume the next person to take their job will not be nearly as valued.

Also, ML will impact programmer demand in subtle ways. A lot of programming is refactoring, and there is reason to believe we can refactor code, especially in certain languages, automatically to make it more aesthetic. Realistically, that seems likely to decrease demand for programmer hours. Or an ML system that can run over someone's GitHub account or repo may be the new resume screen, and if one scores badly on it that may limit the demand for them personally.

Finally, I have to think that the overall march of software towards more complex integrated systems is already a major cause of the dearth of entry-level programming positions, and ML will accelerate that trend.

Take startups. Right now, many startups can compete on the same basis to hire talent as huge companies. But if companies with huge capital reserves can put their cash directly to work to train AI models, startups will be hard-pressed to compete with "smarter" products. Specialization will not even be much help.

Looking at Beating the Averages (http://www.paulgraham.com/avg.html), PG enthused that, since established companies are so behind the curve on software development technology, there is always a chance for higher-productivity techniques like more productive languages to give smaller teams a real chance at a huge market. Of course, that this was in the era when Google was not creating new programming languages and there were no Facebook to widely deploy OCaml and Haskell. And now, AI looks to make the averages even harder to beat.

Even today, if you round up the smartest members of a CS grad class, it is going to be quite difficult to directly compete with a machine learning model with access to huge amounts of data and computing resources. Looking further forwards, if machine learning is able to provide "good enough" alternatives to most human-created software, the software startup narrative — that a few talented and determined people can beat billions in resources — may not even be so relevant anymore.

https://www.scientificamerican.com/article/forearm-gestures-...